Will Jones
By building tissue models from the cells of people who already have Crohn’s disease, VivoSim is extending its lab-grown human tissue platform from predicting harm to predicting whether a drug can help.
Every drug that reaches a pharmacy shelf has cleared several benchmarks to get there. The main two: proving that it is safe enough for the patients who will take it and that it actually treats the condition it was designed for. Most of the attention around lab-grown human tissue models has prioritized the safety bar, because a model that flags liver toxicity or gastrointestinal damage before a drug reaches people can spare patients harm, and spare developers a costly late-stage failure. Research suggests that roughly 90% of drug candidates entering clinical development never receive approval, and those failures can happen at either benchmark.
VivoSim, a San Diego biotech company that grows three-dimensional human tissue to study drug responses before clinical trials, has built much of its work (so far) around safety. The company is now applying the same platform to efficacy, and the shift that makes this possible begins with where the cells originate from.
Why the Donor Matters
When VivoSim tests whether a drug will cause harm, it builds tissue from the cells of healthy donors, exposes that tissue to the compound and watches for damage, such as the intestinal lining beginning to break down in a way that often foreshadows diarrhea and other gastrointestinal side effects in human trials. To test whether a drug can help, the company reverses the setup and grows the same kind of tissue from cells taken from patients who already are affected by the disease.
Keith Murphy, VivoSim’s Executive Chairman says the difference is the starting point of the cells: “The cells come from a healthy person when you want to see if the drug is going to cause a problem, and if you want to see if it’s going to cure, you take the cells from a person with disease.”
Tissue grown from a patient’s cells carries that patient’s disease with it, so the model begins in the same compromised state as the organ it represents, rather than as an idealized healthy version that researchers would otherwise have to damage artificially before testing a treatment.
Inside the Intestinal Model
VivoSim’s intestine model is a thin layer of cells grown on a small membrane, forming a barrier that resembles the lining of the human gut, alongside a thicker layer of cells that make up the intestinal wall. In a healthy intestine, the lining sits tightly along the tissue wall, absorbing nutrients while keeping food particles and bacteria on the other side. In Crohn’s disease and ulcerative colitis, two closely related inflammatory bowel conditions, the lining cells no longer line up tightly next to one another the way they should, and the weakened barrier lets bacteria pass into tissue where they drive inflammation, pain, immune response, and the cascade of symptoms patients live with.
When VivoSim grows the model from cells donated by Crohn’s or ulcerative colitis patients, the tissue develops that same breakdown on its own, without researchers needing to induce it. Developers can then apply a candidate drug and see whether the lining recovers its healthy structure, which gives them a human-based readout of whether the compound is doing what it was designed to do. Because the tissue is grown in a postcard-sized format holding 96 samples at once, teams can compare drug candidates side by side as chemists refine them, tracking whether each chemistry change brings the tissue closer to health. A final drug candidate can be selected from those that do the best job and have been tested by other methods to be suitable as a drug candidate. Outside companies come to VivoSim to test either their compounds at the stage when they have many candidates to screen, or bring final candidates to further characterize how exactly they are working and what patient populations they benefit most.
The company plans to offer the same approach to fatty liver disease using a liver model for MASH (metabolic-associated steatohepatitis), which combines the four major cell types found in the human liver and also takes cells from a patient with that disease.
A Shifting Landscape
According to the company, a team working with VivoSim’s Crohn’s and ulcerative colitis model used the intestinal model with diseased cells to identify a drug candidate for those diseases, which was later sold to Eli Lilly and is now in Phase 2 clinical trials Lilly is conducting.
The timing also lines up with regulatory change. In September, the U.S. Food and Drug Administration finalized a rule that opens the door to non-animal methods, including human cell-based tests, in drug development. Much of the conversation around that rule has centered on safety testing, yet the same reasoning carries over to efficacy, since an animal model of inflammatory bowel disease can only approximate what happens in a human gut, while tissue from patients with the disease starts from the human condition itself.
A patient-derived model will not replace clinical trials, but it offers developers an earlier and more human-relevant signal about whether a drug is worth carrying forward, which is where much of the cost and disappointment in drug development is decided. VivoSim’s bet is that the same tissue that can warn a developer about harm can also show, well before a trial would, whether a drug has a real chance of helping the patients it was built for.